摘要
目的 建立ARIMA模型,对吉林省2024年百日咳发病趋势进行预测,分析ARIMA模型的预测效果,为百日咳防控提供依据。方法 应用Excel 2021和SPSS 23.0软件对2018—2023年吉林省百日咳发病数据进行整理和分析,建立ARIMA模型,并预测2024年吉林省的月发病数。结果 2018—2023年吉林省共报告百日咳病例1 808例,无死亡病例。吉林省百日咳月发病数最佳预测模型为ARIMA(1,1,2)×(0,1,0)_(12)模型。残差检验结果P>0.05,该模型残差为白噪声,具备纯随机性。平稳R^(2)为0.555,R^(2)为0.857,AIC值为549.19,正态化BIC值为7.351,拟合效果良好。该模型预测2024年百日咳发病数呈上升趋势。结论 ARIMA(1,1,2)×(0,1,0)_(12)模型能够较好地拟合吉林省百日咳发病趋势,对吉林省百日咳疫情的预警、预报、防控措施的制定具有一定指导意义。
Objective To establish ARIMA model,predicting the incidence trend of pertussis in Jilin Province in 2024,analyze the predictive effect of the model,and provide a basis for pertussis prevention and control.Methods Excel 2021 and SPSS 23.0 software were used to organize and analyze data on the incidence of pertussis in Jilin Province from 2018 to 2023,establish ARIMA model,and predict the monthly incidence of pertussis in Jilin Province in 2024.Results A total of 1808 cases of pertussis were reported in Jilin Province from 2018 to 2023,with no deaths.The best predictive model for the monthly incidence of pertussis in Jilin Province was the ARIMA(1,1,2)×(0,1,0)_(12) model.The residual test result was P>0.05,indicating that the residual of the model was white noise and has pure randomness.The stationary R2 was 0.555,the R2 was 0.857,the AIC value was 549.19,and the normalized BIC value was 7.351,with good fitting effect.The model predicted an upward trend in the incidence of pertussis in 2024.Conclusion The ARIMA(1,1,2)×(0,1,0)_(12) model can better fit the trend of pertussis incidence in Jilin Province,and has certain guiding significance for the early warning,forecasting,and prevention and control measures of pertussis epidemic in Jilin Province.
作者
曹博楠
付思美
程涛
李春梅
曹凤瑞
武懿
李永强
孙朦
李响
王爽
Cao Bonan;Fu Simei;Cheng Tao;Li Chunmei;Cao Fengrui;Wu Yi;Li Yongqiang;Sun Meng;Li Xiang;Wang Shuang(Jilin Provincial Center for Disease Control and Prevention(Jilin Academy of Preventive Medicine),Changchun,Jilin 130062,China)
出处
《中国公共卫生管理》
2025年第3期438-440,共3页
Chinese Journal of Public Health Management
基金
吉林省卫生健康科技能力提升项目(2023Gw006)。
